With so much hype around AI, the In-House Agency Leaders Club (IHALC) wanted to get under the skin of how in-house creative teams are actually using the technology, what they think the opportunities are with AI and the challenges to fully realising them. So we asked them. Our AI Survey, conducted in partnership with Adobe, analyses responses from in-house agency leaders at over 50 major brands and organisations in the UK and EMEA, across sectors including charities, media, FMCG, financial services and retail.

The potential of AI can no longer be in doubt, but that potential comes with major concerns around ethics, bias, copyright, IP and its impact on the lives of the photographers, illustrators, filmmakers and other specialists on whose skills the creative industry relies. As leaders of in-house teams who, by definition, operate within larger corporate structures, our community is at the sharp end of figuring out how to negotiate this new landscape. Overwhelmingly, they recognise the exciting opportunities of GenAI, but they also recognise their responsibilities to their teams, partners, customers and stakeholders.

AI AND IN-HOUSE AGENCIES: TEST AND LEARN

While the vast majority (78%) of in-house agencies (IHAs) responding to our survey are using generative AI in their operations already, it is very much in the spirit of testing and learning at this stage. Most activity is taking place in creative concepting, with 70% of respondents reporting that they are using GenAI tools to work up ideas, produce multiple options and visualise different routes at speed.

Almost all (93%) had used AI to create images; 62% had used it for body copy and 40% for headlines. Reflecting the amount of adaptation work that most IHAs undertake, 31% had used AI to create voiceovers. In one of IHALC’s recent online sessions, fashion reselling marketplace Vinted told us that it had been experimenting with AI to create local language adaptations across the 21 countries in which it operates.

Moving image usage was less common, with just 20% having used it to create ‘live action’ video and 11% animation – we expect to see these values rise as new tools are launched and existing systems add more sophisticated moving image functionality.

However, relatively few are using GenAI to create assets that have been used in live work for external audiences – just 28%. In a year’s time, respondents told us, this could rise to 50%, but, if GenAI is used in this way, IHAs believe it is important that consumers are made aware of this: when we asked respondents to tell us how important it is that consumers are made aware that images or video have been created using AI on a scale of 1 to 10, 60% gave a score of 8 or more, with 28% scoring it a 10. This underlines the importance of trust and transparency when it comes to AI.

COPYRIGHT AND IP ARE THE BIGGEST CONCERNS

While the technology exists to enable brands to use GenAI to create their campaigns and content, legal and regulatory concerns mean that, in reality, organisations are more cautious about its use.

Our survey revealed that 58% of organisations have AI policies in place (AI policies or strategies specifically for the IHA are less prevalent, with only 28% of respondents reporting that they had one).

These policies typically cover governance, compliance, legal and security issues, with an emphasis on protecting the organisation from exposure to legal action and keeping its own IP and data safe. Some organisations reported a blanket ban on the use of any AI tools, while others cited lengthy corporate approvals processes before anything can be deployed by IHAs.

Copyright infringement and IP safety were cited as the biggest challenges by respondents, but ethics issues are also being considered, with leaders expressing concerns that AI systems treat artists with fairness.

Creatives also cited their own ethical concerns about what AI will mean for the craft industries and that overuse may “hollow out” creativity and lead to “Boring, eye-candy work that is devoid of meaning and therefore forgettable”. The challenge will be to maintain distinctiveness.

Sustainability issues were also raised, particularly for organisations with environmental targets to hit.

AUTOMATING THE MUNDANE TO SET US FREE

Notwithstanding their concerns about AI, IHAs are generally optimistic and enthusiastic about its potential. This optimism also translates to their teams, with our data showing a broadly positive attitude to AI.

Rather than seeing AI as a threat to jobs (only 22% thought the introduction of AI would result in them having a smaller team), the overriding view among respondents is that ‘AI’s role in creativity will be primarily as an assistant to human creative teams’ (80% of respondents agreed with this statement). This is a vision of GenAI as a co-pilot, not a replacement.

By automating mundane tasks that soak up so much of their time currently, the majority of respondents believe AI will enable in-house teams to concentrate on higher-value work. As one respondent told us, “As a team, we’ve always been busy and are always under-resourced. Automating mundane tasks is a great opportunity for us. It will free everyone up from the tedious work, allowing them to focus on more strategic, high-value activities that drive real impact and innovation. This not only benefits the company but also allows everyone to tap into their full potential and find more creative solutions.”

When we asked our respondents to score the potential of AI across a number of opportunities, the highest scorers (ie the areas in which our respondents feel that AI will have the greatest positive impact on their work), were improving briefs, creative concepting, automating mundane tasks (81% scoring it an 8 or higher), and saving time (74% scoring 8 or higher).

When it came to cost saving, only 50% scored AI an 8 or higher, with respondents perhaps mindful of the costs of setting up AI systems and the potential gap between their promised impact and reality. Respondents were similarly unenthused about AI’s potential to lead to new styles or genres of work. Over 50% scored AI at 5 or less for this.

FROM PLAYGROUND TO PRODUCTION

Our survey suggests that most IHAs are on the cusp of transitioning AI from experimentation to deployment. But there are significant issues to work through.

Organisations are understandably wary of the copyright and IP implications. In some cases this has led to an outright ban on the use of AI systems. Others have created ethics committees, often comprising representatives from legal, HR, procurement, marketing and the IHA, to determine and review their approach, including which tools they will use and on what.

IHA leaders believe it is important that consumers are made aware AI has been used to create campaign assets. How that will be done, however, is unclear. One route is via the Content Authenticity Initiative, announced by Adobe in 2019, which includes the BBC, Getty Images, Intel, Microsoft and the New York Times among its members. It proposes using an online verification system based on metadata.

It seems highly likely that the increased use of ‘closed’ AI systems (our survey suggests that just 28% of IHAs are using these currently will alleviate some concerns around IP and copyright. Brands will be able to train systems using their own data and images and ensure that none of the content created by those systems makes its way out into the public versions of those tools. But that will require brands to have usage rights for those assets and for the assets to be accessible.

What comes through strongly in the survey is the hope that AI will allow in-house teams to ‘automate the mundane’ and that this will free them up to focus on higher value creative solutions to business problems. While the majority believe that this will not result in them having smaller teams, the impact of fully deployed AI systems, particularly on roles that are currently responsible for lower-tier work, may make that view seem optimistic.

At the very least, it seems likely that roles will change and new skills will need to be acquired, just as they were when desktop publishing impacted old-school print production. It will fall to IHA leaders to exert their influence to ensure that AI results in better work, not just cheaper and faster work. But as the latter is more easily quantifiable, this may be a hard line to maintain.

While the focus of our survey is on the remit of IHAs themselves, we are also aware of brands deploying AI across the wider organisation with astonishing results. This is particularly true of new product development, where some brands have conducted tests asking AI to come up with brand extensions, getting back detailed results in minutes that go all the way from market analysis to naming, to pack shots. The impact of AI will be felt upstream in the marketing department, as well as downstream in content production and delivery.

Finally, we should remember that, as with previous technological revolutions, AI is about to go through a ‘bedding in’ stage, where the hype meets the real world. As one of our respondents said, “My hope is that, with AI, we’re at the ‘rotating logos on fire’ moment, like we were when the internet was the new shiny thing. We played with that for a long time, doing things just because we could, and then it settled down and we found the best way to harness its power. I hope AI will be the same story.”

Download the full report here; ihalc.com